IntroVAE: Introspective Variational Autoencoders for Photographic Image Synthesis

07/17/2018
by   Huaibo Huang, et al.
10

We present a novel introspective variational autoencoder (IntroVAE) model for synthesizing high-resolution photographic images. IntroVAE is capable of self-evaluating the quality of its generated samples and improving itself accordingly. Its inference and generator models are jointly trained in an introspective way. On one hand, the generator is required to reconstruct the input images from the noisy outputs of the inference model as normal VAEs. On the other hand, the inference model is encouraged to classify between the generated and real samples while the generator tries to fool it as GANs. These two famous generative frameworks are integrated in a simple yet efficient single-stream architecture that can be trained in a single stage. IntroVAE preserves the advantages of VAEs, such as stable training and nice latent manifold. Unlike most other hybrid models of VAEs and GANs, IntroVAE requires no extra discriminators, because the inference model itself serves as a discriminator to distinguish between the generated and real samples. Experiments demonstrate that our method produces high-resolution photo-realistic images (e.g., CELEBA images at 1024^2), which are comparable to or better than the state-of-the-art GANs.

READ FULL TEXT

page 7

page 8

page 9

page 14

page 15

page 16

research
09/09/2019

An Acceleration Framework for High Resolution Image Synthesis

Synthesis of high resolution images using Generative Adversarial Network...
research
05/31/2018

On GANs and GMMs

A longstanding problem in machine learning is to find unsupervised metho...
research
09/11/2016

Energy-based Generative Adversarial Network

We introduce the "Energy-based Generative Adversarial Network" model (EB...
research
10/30/2016

Conditional Image Synthesis With Auxiliary Classifier GANs

Synthesizing high resolution photorealistic images has been a long-stand...
research
07/09/2018

Pioneer Networks: Progressively Growing Generative Autoencoder

We introduce a novel generative autoencoder network model that learns to...
research
06/15/2020

Generalized Adversarially Learned Inference

Allowing effective inference of latent vectors while training GANs can g...
research
04/01/2020

Manifold-Aware CycleGAN for High Resolution Structural-to-DTI Synthesis

Unpaired image-to-image translation has been applied successfully to nat...

Please sign up or login with your details

Forgot password? Click here to reset